Music training and the brain

Transcription

1 Why we study the neuroscience of music and other art forms as a window to the creating brain Fredrik Ullén, Dept of Neuroscience, Karolinska Institutet East-West Connections, Singapore, 2016 Intrinsically interesting neuroaesthetics Interventions Model behaviors for cognition in general Sensory discrimination Motor performance and coordination Emotion Social cognition Executive control Skill learning and expertise Creativity etc Why we study the neuroscience of music and other art forms Intrinsically interesting neuroaesthetics Interventions Model behaviors for cognition in general Sensory discrimination Motor performance and coordination Emotion Social cognition Executive control Skill learning and expertise Creativity etc training and the brain Training and expert performance Gray matter in musicians Krampe and Ericsson (1996), J Exp Psychol: General Gaser and Schlaug (2003), J Neurosci 1

2 Specific gray matter effects of different musical training al training and gray matter longitudinal data Omega Sign (OS) in precentral gyrus OS-1 or OS-2 OS-2 Bangert and Schlaug (2006), Eur J Neurosci Hyde et al (2009), J Neurosci Regional gray matter structure and performance White matter in musicians Schneider et al (2002), Nat Neurosci Bengtsson et al (2005), Nat Neurosci White matter in musicians More well-organized corpus callosum in early trained musicians al training and white matter longitudinal data Steele et al (2013), J Neurosci Hyde et al (2009), J Neurosci 2

3 Summary al training is correlated with brain anatomy Gray matter (regional volume, cortical thickness) White matter (organization, connectivity) Neuroanatomical effects are specific to type of musical training -related Longitudinal data show differences developing over time in trainers versus non-trainers Training effects the problem of causality The problem of causality 1 The problem of causality 2 Cross-sectional data Many causal scenarios are possible Observational longitudinal studies Scan 1 Scan 2? group? Control group Time The problem of causality 3 Randomized longitudinal studies Ideal in principle Practically difficult/impossible in expertise studies Months/years of dedicated Case in point: music training and musical ear al ear = musical auditory discrimination Ability to discriminate rhythms, melodies, pitches etc Positively related to musical training ians outperform non-musicians Commonly assessed in entrance exams to music colleges 3

4 Humans making music Web-based data collection (finished feb 2013) - responses from > 10,500 twin individuals training al childhood environment al auditory discrimination, reaction time Creative achievement (7 domains) General personality (Big Five), schizotypy Intrinsic and extrinsic motivation (GSM) Occupational preferences Proneness for flow experiences General psychological and somatic health Emotional processing (alexithymia) What drives the association between training and musical ear? Participants 1211 monozygotic ( identical ) pairs (MZ) 1358 dizygotic pairs (DZ) 5401 single twin individuals Measures Total hours of musical Swedish Discrimination Test (SMDT) Melodies Rhythms Pitches al training and music discrimination Melody Rhythm Pitch Training Melody Rhythm % of variance in ability explained by training other things matter! Twin modelling Classical twin design Compare MZ and DZ twins Partition phenotypic variance and covariance into: A: Additive genetic effects C: Shared environment E: Non-shared environment Genetic factors influence training and musical abilities al training and music discrimination ability Melody Rhythm Pitch Heritability Training Melody Rhythm Pitch.39 No significant within-pair relations for any of the music discrimination scales (r values ) 4

5 What drives the association between training and musical ear?! Practice and expert performance - metaanalyses only explains a moderate proportion of variance in expert performance? Macnamara et al (2014), Psychol Sci 5

6 Genetic influences on musical discrimination increase with training Variance component A C E training hours (SDs from mean) Verweij, Mosing et al (under review) How can we get at causal effects of long-term training in expertise? Monozygotic twins intrapair difference model If training causes an outcome, we would expect Difference in training (twin 1 twin 2) Difference in outcome (twin 1 twin 2) Humans making music Web-based data collection (finished feb 2013) - responses from > 10,500 twin individuals al training al childhood environment al perceptual abilities (pitch, rhythm, melody), reaction time Creative achievement (7 domains) General personality (Big Five), schizotypy Intrinsic and extrinsic motivation (GSM) Occupational preferences Proneness for flow experiences General psychological and somatic health Emotional processing (alexithymia) Experimental studies on smaller sample of extremely discordant MZs Monozygotic twins discordant for piano 10 pairs with > 1300 h intrapair difference in piano (recruited from all over Sweden) playing twin : still active non-playing twin : no in adulthood all right-handed 6

7 Experiments Interviews Working memory tests Finger force control tests MR scanning Brain anatomy (structural MRI) Gray matter, diffusion MRI Functional MRI Sequence production, perception, improvisation Why did the twins differ in musical engagement? Psychological questionnaire data Playing twin higher in Enjoying music Playing twin more frequent psychological flow experiences during musical activities Eriksson et al (in press), ae Scientiae Why did the twins differ in musical engagement? - Interviews Semi-structured interviews with five main themes Own thoughts about why they became discordant Childhood differences in musical environment music listening, music teachers, music interest of peers, parental engagement, public performing Strong memories of music Significance of music in your life Interest and in language Why did the twins differ in musical engagement? - Interviews Playing twin More elaborate answers about meaning of music in life, emphasizing importance for personal identity No systematic within-pair differences in self-reported interests of peers, parental support, music teacher, ensemble playing, public performances, interest and aptitude for languages Eriksson et al (in press), ae Scientiae Eriksson et al (in press), ae Scientiae Why did the twins differ in musical engagement? - Interviews Reported possible reasons for discordance idiosyncratic and unique for each pair, e.g.: Different access to the piano Different feelings about music genre played at home Different needs for creative hobbies Different feelings about the music teacher Different attitudes to music as an expression of faith -playing parent role model for one but not other twin etc When controlling for genes and shared environment, remaining influences are unsystematic in nature Summary Discordant monozygotic twins provide a unique opportunity to study pure training effects (i.e. independent of genetic factors and common environment) Trained twin Larger regional gray matter volume of left temporo-parietal junction Higher fractional anisotropy in left pyramidal tract of playing twin Higher working memory capacity for musical materials Eriksson et al (in press), ae Scientiae 7

8 Improvisation as a model for creative performance as a window to the creating brain Ecologically valid Possible to study with brain imaging (simplified!) What is the role of the DLPFC for improvisation? Dorsolateral prefrontal cortex (DLPFC) Attention, working memory, selection The DLPFC is activated when classical pianists improvise Improvisation (ornaments) versus from memory (Bengtsson et al, 2007; J Cogn Neurosci) L y = 39 R DLPFC t-values STG x = -60 y = 6 L z = 45 PMD PMD R Improvisation versus playing from score (de Manzano et al, 2012; NeuroImage) The DLPFC is deactivated during jazz improvisation in jazz pianists Improvisation training and the brain Is specific jazz expertise one factor behind these findings? Does training improvisation have specific effects over and above classical piano training? 39 pianists (varied jazz/classical background) Brain activity during brief improvisations Associations with improvisation training (controlling for classical training) Limb and Braun (2008), PLoS One Pinho et al (2014), J Neurosci 8

9 Improvisation training negatively correlated with DLPFC activity during improvisation DLPFC Summary The involvement of the DLPFC in improvisation is lower for experienced improvisers Characteristics of the task also matter! (Pinho et al, 2015, Cereb Cortex) Improvise in a given mood low DLPFC Improvise using a given pitch set high DLPFC Pinho et al (2014), J Neurosci Summary The involvement of the DLPFC in improvisation is lower for experienced improvisers Characteristics of the task also matter! (Pinho et al, 2015, Cereb Cortex) Improvise in a given mood low DLPFC Improvise using a given pitch set high DLPFC Different strategies for creative thinking? Low top-down control - high level of task-specific expertise, allow spontaneous free association High top-down control lower level of task-specific expertise, free association not useful General summary Artistic behaviors and professional artists are unique models for higher human cognition and expertise is multifactorial and depends on geneenvironment interactions Twins are useful for studies of causal effects of long-term training and creativity Improvisation as a model for creative cognition Prefrontal involvement in creative thinking depends on training and task characteristics Coworkers & collaborations Post-docs and alumni Miriam Mosing, Karin Verweij, Örjan de Manzano Helene Eriksson, László Harmat, Anders af Wåhlberg PhD students Ana Luísa Pinho, Diana Müssgens, Lars Forsberg IT, administration Pelle Karlsson, Louise von Essen Collaborations Nancy Pedersen, Töres Theorell (KI), Guy Madison (Umeå University) Zach Hambrick (Michigan State Univ), Mihály Csíkszentmihályi (Claremont Grad Univ), Mark Hallett (NIH), Miguel Castelo-Branco (University of Coimbra) Funding Bank of Sweden Tercentenary Fund Sven and Dagmar Salén Foundation Swedish Brain Foundation Kjell and Märta Beijer Foundation Swedish Scientific Council 9